Distributed User Association and Computation Offloading in UAV-Assisted Mobile Edge Computing Systems

Unmanned Aerial Vehicle (UAV)-assisted Mobile Edge Computing (MEC) systems have emerged as promising solutions for enhancing the computational capabilities and reducing latency in next-generation wireless networks. However, the finite energy capacity of UAVs presents a significant challenge. In this study, we formulate a joint user association, computing resource allocation, and task offloading problem, called User Association and Computation Offloading (UACO), to minimize the energy consumption of both mobile devices and UAVs by considering the computational resource limitations of UAVs and the minimum user offloading data rate requirements. The UACO problem is a mixed-integer linear programming (MILP) problem that makes it NP-hard. We transform UACO into an Offload Exact Constrained Potential Game (OFECPG) based on game theory, which facilitates distributed execution. We propose the Best Response dynamics based on Local Decomposition (BR-LD) and Better Response dynamics based on the Local Switch operator (BR-LS) to enhance computational efficiency. We prove the existence of a pure strategy Nash Equilibrium (NE) and the convergence of the proposed algorithms. Extensive simulations demonstrated the effectiveness of the OFECPG, BR-LD, and BR-LS, showing significant improvements in system energy consumption compared to the baseline schemes. Our approach offers valuable insights into the design of efficient association and offloading schemes for multi-UAV MEC networks.